Toxicology of ZnO and TiO2 nanoparticles on hepatocytes: impact on metabolism and bioenergetics.

http://informahealthcare.com/nan ISSN: 1743-5390 (print), 1743-5404 (electronic) Nanotoxicology, Early Online: 1–9 ! 2014 Informa UK Ltd. DOI: 10.3109/17435390.2014.895437

ORIGINAL ARTICLE

Nanotoxicology Downloaded from informahealthcare.com by University Of Rochester School Of Medici ne & Dentistry on 06/02/14 For personal use only.

Toxicology of ZnO and TiO2 nanoparticles on hepatocytes: Impact on metabolism and bioenergetics Celine Filippi1,2*, Anne Pryde1, Pauline Cowan1, Tricia Lee1, Peter Hayes1, Ken Donaldson3, John Plevris1, and Vicki Stone2 1

Hepatology Lab, University of Edinburgh, Chancellor’s Building, Edinburgh, UK, 2Nanosafety Research Group, Heriot-Watt University, Edinburgh, UK, and 3ELEGI-COLT Lab, Centre for Inflammation Research, Edinburgh, UK

Abstract

Keywords

Background and aim: Zinc oxide (ZnO) and titanium dioxide (TiO2) nanomaterials (NMs) are used in many consumer products, including foodstuffs. Ingested and inhaled NM can reach the liver. Whilst their effects on inflammation, cytotoxicity, genotoxicity and mitochondrial function have been explored, no work has been reported on their impact on liver intermediary metabolism. Our aim was to assess the effects of sub-lethal doses of these materials on hepatocyte intermediary metabolism. Material and methods: After characterisation, ZnO and TiO2 NM were used to treat C3A cells for 4 hours at concentrations ranging between 0 and 10 lg/cm2, well below their EC50, before the assessment of (i) glucose production and glycolysis from endogenous glycogen and (ii) gluconeogenesis and glycolysis from lactate and pyruvate (LP). Mitochondrial membrane potential was assessed using JC-10 after 0–40 lg/cm2 ZnO. qRT-PCR was used to assess phosphoenolpyruvate carboxykinase (PEPCK) mRNA expression. Dihydroethidium (DHE) staining and FACS were used to assess intracellular reactive oxygen species (ROS) concentration. Results: Treatment of cells with ZnO, but not TiO2, depressed mitochondrial membrane potential, leading to a dose-dependent increase in glycogen breakdown by up to 430%, with an increase of both glycolysis and glucose release. Interestingly, gluconeogenesis from LP was also increased, up to 10-fold and correlated with a 420% increase in the PEPCK mRNA expression, the enzyme controlling gluconeogenesis from LP. An intracellular increase of ROS production after ZnO treatment could explain these effects. Conclusion: At sub-lethal concentrations, ZnO nanoparticles dramatically increased both gluconeogenesis and glycogenolysis, which warrants further in vivo studies.

Gluconeogenesis, mitochondria, nanomaterial, toxicology

Introduction Nanomaterials (NMs) are defined as having one or more dimensions in the 1–100 nm range. Decreasing the size of bulk particles down to nanometre size confers them novel properties, which allowed diverse and totally new applications. Hundreds of thousands of tons of zinc oxide (ZnO) and titanium dioxide (TiO2) nanoparticles are manufactured each year for use in paints, coatings, tyre rubber and liquid crystal displays, to name but a few (Klingshirn, 2007). TiO2 and ZnO also show photocatalytic and biocidal activity (Brayner et al., 2006; Jones et al., 2008), justifying their use in self-cleaning windows, water filtration systems or materials for public buildings where hygiene is primordial. A major concern for toxicologists, beside the potential increase in exposure of industry workers, is the use of NMs for the production of every-day consumer products such as

*Present address: NIHR Biomedical Research Centre at Guy’s and St Thomas’ NHS Foundation Trust and King’s College London, Institute of Liver Studies, King’s College London, Bessemer Road, London, UK Correspondence: Dr. Celine Filippi. Tel: +44 203 299 1545. Fax: +44 203 299 3760. E-mail: [email protected]

History Received 6 April 2013 Revised 10 February 2014 Accepted 12 February 2014 Published online 7 April 2014

transparent UV protection, toothpaste or even food colouring (Nanotoxicology consumer products, 2012; Weir et al., 2012). This has led to the development of nanotoxicology to investigate exposure, hazard and risk potential. Due to their small size, NMs have an increased propensity to be adsorbed and cross biological barriers, which may pose an extraordinary hazard to organisms (Miller et al., 2010; Stone & Donaldson, 2006). In experimental studies, NM delivered into the lungs or the stomach (Jani et al., 1990; Sycheva et al., 2011) become blood borne and reach secondary target organs such as the brain, spleen, kidneys and the liver (Donaldson et al., 2005; Oberdo¨rster et al., 2002). Particles interacting with these organs can deliver an oxidative stress to the cells (Donaldson et al., 2005; Hussain et al., 2005; Xia et al., 2004) and trigger the same increase in acute phase proteins as seen in populations exposed to increased ambient air fine particulate matter (Monn & Becker, 1999; Stone et al., 2007). Exposure to ultrafine particulate matter has already been shown to activate Kupffer cells and exacerbate non-alcoholic fatty liver disease (NAFLD) in animal models (Tan et al., 2009). Furthermore, it is also known that increases in such liver-derived mediators as interleukin-6, C-reactive protein and fibrinogen are risk factors for death from acute coronary syndrome (Koukkunen et al., 2001).


2

C. Filippi et al.

Most studies undertaken so far, in vivo or in vitro, dealt with inflammation (Khandoga et al., 2004; Monn & Becker, 1999; Xia et al., 2008), oxidative damage (Kamat et al., 1998; Pujalte´ et al., 2011; Saquib et al., 2012; Sharma et al., 2011, 2012; Sun et al., 2011), cytotoxicity (Hussain et al., 2005; Pujalte´ et al., 2011; Saquib et al., 2012; Sharma et al., 2012; Wang et al., 2011) or DNA damage (Saquib et al., 2012; Sharma et al., 2011) induced by NMs. Few studies have focused on mitochondrial function itself (Hussain et al., 2005; Sun et al., 2011; Xia et al., 2004) and, to our knowledge, none have dealt with liver metabolism. In the clinic, there is a relatively large proportion of cirrhosis cases of unknown causes, described as cryptogenic – 5–30% depending on the studies (Maheshwari et al., 2006). Mitochondrial dysfunction is one of the triggers of NAFLD leading to liver metabolic disturbances and a wide range of damage, with cirrhosis at the end of the spectrum. Our aim was therefore to analyse the effects of ZnO and TiO2 nanoparticles on hepatocyte mitochondrial function and intermediary metabolism, at concentrations known not to affect cell viability. Our study not only confirms the impact of the ZnO nanoparticles on hepatocyte mitochondrial function but also shows a profound modification of gluconeogenesis regulation linked to over-expression of its main regulator, the phosphoenolpyruvate carboxykinase (PEPCK). TiO2, on the other hand, had limited effects, thus supporting previous reports describing it as one of the least toxic NMs (Kermanizadeh et al., 2012).

Nanotoxicology, Early Online: 1–9

endogenous substrates only or (ii) using a mix of physiological substrates (lactate 20 mM, pyruvate 2 mM, octanoate 4 mM and ammonium chloride 4 mM). The measurement of lactate (L), pyruvate (P) and glucose (G) in the cell supernatant according to Bergmeyer et al. (1974) allowed us to assess the glycolysis flux as the sum of the L and P fluxes (JL + P), the cytosolic redox potential, assessed by the ratio of L to P (L/P) (Williamson et al., 1967) and the glucose production flux (Jglc), which results either from glycogen breakdown (for the endogenous substrate condition) or gluconeogenesis (from lactate and pyruvate; LP). In physiological conditions, the balance of glycogen breakdown towards glycolysis or gluconeogenesis is partly controlled by the availability in ATP (Constantin et al., 1995), whilst the gluconeogenesis from LP is regulated by the level of expression of the PEPCK (Groen et al., 1983). Effect of NM on mitochondrial membrane potential (DWm) After exposure to NM, the cells were rinsed with HBSS and stained with JC-10 (1 mM, 20 min at 37 C). In active mitochondria J-aggregates of JC-10 form, resulting in a shift of peak fluorescence from 520 to 570 nm, measured in a plate reader. Because of the dramatic effects of ZnO on this measurement, this particular set of experiments included smaller concentrations of NM, obtained by serial dilutions. Controls with unstained cells did not show any significant signal due to the NM itself. RNA isolation and qRT-PCR

Methods Characterization of NM The surface areas of ZnO (10.7 ± 0.7 nm, NanoScale Corporation, Manhattan, KS) and rutile TiO2 (30.5 ± 1.8 nm, Nanostructure and Amorphous Materials Inc., Houston, TX) were determined with a Micromeritics TriStar3000 (Bedfordshire, UK) by Escubed Ltd. (Leeds, UK) after dispersion in saline and 5% heat-inactivated serum. Primary particle sizes were assessed by transmission electron microscopy. Hydrodynamic sizes were assessed with a Brookhaven 90 plus (Holtsville, NY). The zeta potentials were measured using a Zetasizer-Nano ZS instrument (Malvern, Malvern Hills, UK), in cell culture medium or an artificial lysosomal fluid medium with a pH of 5.5. The capacity of the particles to produce reactive oxygen species (ROS) was measured by electro-paramagnetic resonance (EPR) (Miller et al., 2009). Effects of NM on liver intermediary metabolism The C3A cell line (ATCCÕ CRL-10741Ô), a clonal derivative of the hepatocarcinoma HepG2 cell line, selected for its higher state of differentiation, was cultured in six-well plates (10 cm2 per well) in 3 ml Minimum Essential Medium Eagle (MEME; Sigma Aldrich, Dorset, UK) with 10% foetal calf serum and Pen/Strep (Sigma) until 80% confluent, before exposure to nanoparticulate TiO2 or ZnO. Each NM was first suspended at 1 mg/ml in phosphate buffered saline (Sigma) and sonicated in a waterbath sonicator without pause for 16 min at 25 C before addition to the cells at final densities of 0–10 mg/cm2, in a volume of 1 ml of full culture medium, for 4 hours. This range of low concentrations was chosen to be well below the LC50 and to resemble potential hepatocyte exposure doses that could occur in vivo after repeated inhalation or ingestion of NM (Kermanizadeh et al., 2013; Kreyling et al., 2009). The exposure time of 4 hours was chosen to detect early events of NM exposure, rather than chronic effects. After exposure, the cells were rinsed with phosphate buffer saline (Life Technologies, Paisley, UK) before being incubated for 4 hours in HBSS with calcium and magnesium (i) without any exogenous addition to test the metabolism from

RNA was isolated from C3A cells using a MagMax 96 total RNA isolation kit (Ambion/Life Technologies, Paisley, UK). After RNA quality and concentration analysis using a Nanodrop, a one-step qRT-PCR (Primer Design kit, Southampton, UK) was performed to detect the PEPCK (PCK1) gene expression level (ABI Hs00159918_m1), normalised to the b2m housekeeping gene expression (Primer Design). ZnO dissolution and release of Zn2+ To test the solubility of ZnO and subsequent release of Zn2+ ions, ZnO NM was incubated in MEME, for 4 hours and at the same concentrations as during the cell exposure (100, 25 or 5 lg/ml in 10 cm2 cell culture wells, corresponding to 10, 2.5 or 0.5 lg/cm2). The culture medium was then shaken vigorously to make sure any dissolved Zn2+ is homogenously distributed, before being filtered using Vivaspin ultrafiltration spin columns, with a cut-off size of 5000 Da, using a 20 min centrifugation at 5000g. The concentration of Zn2+ ions in the filtrate was analysed using air/acetylene atomic absorption spectroscopy (PerkinElmer, Cambridge, UK). Intracellular ROS measurements After exposure to NM, C3A cells were trypsinised into a single cell suspension and stained with DHE (2 lM, Life Technologies, Paisley, UK) in HBSS, for 30 min at 37 C. Cell fluorescence was analysed using a FACS-Array (Becton Dickinson, Oxford, UK). H2O2 (5 mM) or silver NM (NM300 from RAS GmbH; Ag capped with polyoxylaurat Tween-205 20 nm at 10 lg/cm2) exposure were used as positive controls. Controls with unstained cells did not show any significant shift in signal due to potential interaction of intracellular NMs with the fluorescence measurement. Statistics GraphPad Prism 4 (Graphpad, La Jolla, CA) was used for graphs and statistical analysis. Two-way analysis of variance tests were used with Bonferroni post-test. Data are presented as mean ± SEM. Each experiment was performed in triplicates for at least n ¼ 3.

Impact of ZnO and TiO2 on hepatocyte metabolism and bioenergetics

DOI: 10.3109/17435390.2014.895437

cells to TiO2, at concentrations up to 10 lg/cm2, did not affect LP metabolism.

Results NM characterisation


3

NM impact does not depend only on the size of the particles but also on their surface area, their zeta potential, their reactivity and the way these features are modified in various aqueous environments, including their capacity to dissolve (Cho et al., 2011; Oberdo¨rster et al., 2005; Xia et al., 2008). Table 1 gives the characteristics of the ZnO and TiO2 particles used for this study. The zeta potential was higher than –30 mV suggesting that the particles would possess a tendency to agglomerate in full MEME medium. The TiO2 particles generated ROS in the cell-free EPR assay suggesting significant surface reactivity, which was not measured for ZnO.

ZnO solubility Part of the toxicity of ZnO has been shown to be due to its solubility, and therefore the production of Zn2+ ions into the cell culture medium or intracellular compartments (Reed et al., 2012; Xia et al., 2008), we therefore checked the solubility of our own particles. The concentrations of dissolved Zn2+ in the 100 lg/ml ZnO particle samples – corresponding to our highest exposure of 10 lg/cm2 on the cells – was 3.9 ± 0.4 lg/ml, indicating that the particle solubility was only 4%, at this time point, 4 hours, and in the conditions used to expose the cells. PEPCK mRNA expression

Effect of ZnO and TiO2 on hepatocyte glycogen metabolism in vitro ZnO induced a striking dose-dependent increase in glycogen breakdown (2G + L + P) in liver cells – up to a 430% increase at 2.5 mg/cm2 (Figure 1; p50.0001). The increase in glycogen breakdown was linked to a large increase in both glycolysis and glucose production. The glycolysis, JL + P, went from 77 ± 13 to 348 ± 44 mmol.gTP1.hr1 (p50.0001) for the highest concentrations of ZnO, a 450% increase. This did not result in a statistically significant increase in the cytosolic redox potential (L/P) (p ¼ 0.79). The glucose flux, JGlc, increased from 8 ± 3 to 46 ± 14 mmol.gTP1.hr1 for 2.5 mg/cm2 of ZnO (p ¼ 0.0003). In contrast, TiO2 showed a very limited effect on glycogen breakdown, glucose, LP release, even at the highest doses tested. Effect of ZnO and TiO2 on hepatocyte DWm Figure 1(E and F) show that the lower doses of ZnO (0.08 lg/cm2) resulted in a dramatic drop of the DWm, to values similar to those obtained with 50 lM of the classical mitochondrial uncoupler carbonyl cyanide m-chlorophenyl hydrazine (CCCP). NM effect on gluconeogenesis To further analyse the effects of NM on glucose metabolism, we next studied gluconeogenesis from LP. Gluconeogenesis from LP is regulated by the PEPCK expression level (Rognstad, 1979). However, any decrease in the production of ATP strongly downregulates glucose production from LP (Leverve et al., 1994). It was therefore surprising to observe a 20-fold increase in gluconeogenesis (JGlc) from LP after exposure to ZnO (101 ± 39 lmol.hr1.gTP1 at 10 lg/cm2 ZnO vs 4.8 ± 0.8 for the controls; Figure 2), when mitochondrial oxidative phosphorylation was affected (Figure 1E). Conversely, exposure of C3A

As stated above, long-term regulation of gluconeogenesis from LP is controlled by the expression level of the PEPCK. We therefore checked the mRNA expression levels of PEPCK in C3As exposed to both NM. To test the potential effects of soluble Zn ions on PEPCK expression, we also analysed the effects ZnCl2 exposure, with concentrations matching that of the ZnO particle exposure (Figure 2). The exposure of C3A cells to ZnO nanoparticles increased the mRNA expression of the PEPCK, in a dose-dependent fashion, up to an approximate four-fold at the 10 lg/cm2 dose (p50.0001). Soluble Zn2+ ions also increased PEPCK mRNA expression. ZnO and TiO2 – ROS production Oxidative stress has been shown to upregulate the transcription of the PEPCK gene in liver cells (Ito et al., 2006; Sutherland et al., 1997). Furthermore, intracellular ROS induce mitochondrial oxidative phosphorylation dysfunction (Lee et al., 2005), which could explain the effects of ZnO on DWm. Therefore, we next assessed (i) the capacity of ZnO and TiO2 to produce ROS in a cell-free system and (ii) the capacity of the NM to induce intracellular ROS production. Indeed, the cell-free system does not take into account the potentially different reactivity of the particles inside the cells (Li et al., 2012; Sharma et al., 2011) or consider the effects of the nanoparticles on the function of the mitochondria and cytochrome P450-dependent oxygenases, which physiologically produce ROS in the cell (Turrens, 2003). The ZnO NM used in our cell experiments displayed a relatively small production of ROS in a cell-free environment (see Table 1, EPR intensity of 619 ± 23), whilst the TiO2 particles showed an EPR intensity of 2004 ± 80 (p50.001 as compared to the control). However, DHE and FACS analysis of intracellular ROS production (Figure 3) showed that ZnO induced a dosedependent intracellular increase of ROS production, with higher

Table 1. Physicochemical characteristics of TiO2 and ZnO NM. Zeta potential (mV) Primary size (nm) TiO2 ZnO a

30.5 ± 1.8 10.7 ± 0.7

Hydrodynamic size (nm) 119 ± 16 306 ± 42

2

1

Surface area (m g ) 27.5 48.2

a

EPR intensity b

2004 ± 80 619 ± 23

Acid

Serum

9.2 ± 0.1 5.5 ± 0.2

–5.6 ± 0.3 –20.5 ± 0.6

Vehicle control: 246 ± 3.8 and positive control (100 mM of pyrogallol): 9630 ± 807. p 5 0.001 vs control. The surface areas of ZnO (10.7 ± 0.7 nm, NanoScale Corporation, Manhattan, KS) and rutile TiO2 (30.5 ± 1.8 nm, Nanostructure and Amorphous Materials Inc.) were determined with a Micromeritics TriStar3000 (Bedfordshire, UK) by Escubed Ltd. (Leeds, UK), after dispersion in saline and 5% heat-inactivated serum. Primary particle sizes were assessed by transmission electron microscopy. Hydrodynamic sizes were assessed with a Brookhaven 90 plus (Holtsville, NY). The zeta potentials were measured using a Zetasizer-Nano ZS instrument (Malvern, Malvern Hills, UK), in cell culture medium or an artificial lysosomal fluid medium with pH5.5 to reflect the situation in intracellular organelles. The capacity of the particles to produce reactive oxygen species was measured by electro-paramagnetic resonance.

b


4

C. Filippi et al.


Figure 1. Intermediary metabolism and mitochondrial membrane potential from endogenous substrates in C3A cells exposed to ZnO or TiO2 nanomaterial. C3A cells were submitted to nanomaterials (0–10 lg/cm2), for 4 hours, before metabolism measurements from endogenous substrates. (A) Glycogen breakdown; (B) Glucose production; (C): Glycolysis; and (D): Cytosolic redox potential. For DWm assessment of mitochondrial function, the cells were submitted to 0–40 lg/cm2 ZnO or 0–100 lM of CCCP, a classical mitochondrial uncoupler (E and F). *p50.0001 vs TiO2.

levels than TiO2 (13.7 ± 0.8 vs 4.5 ± 1.05% of DHE-positive cells, p50.0001), thus potentially explaining the increased PEPCK expression and mitochondrial dysfunction. Table 2 gives as summary of all metabolism data comparison for the two NM studied in this work.

Discussion ZnO effect on hepatocyte intermediary metabolism and mitochondrial function All the parameters measured in our study – the large drop in DWm, and also the increased glycogen breakdown and glycolysis – indicate that ZnO incubation strongly affects C3A cell bioenergetics. ZnO increased glycogen breakdown, glycolysis and glucose release. Gluconeogenesis from LP was also dramatically

upregulated, which correlated with an increase in PEPCK mRNA, the enzyme controlling gluconeogenesis from LP. The increase of intracellular ROS production after ZnO treatment could explain these effects. Remarkably, such changes are similar to the events leading to NAFLD (Pessayre, 2007). No such effects were observed for TiO2 despite an inherent ability to generate ROS in a cell free environment. Glycogen breakdown is regulated by a complex interplay of protein phosphorylation/dephosphorylation, which is controlled by the insulin/glucagon ratio, catecholamines and the intracellular ATP concentration; any decrease in ATP results in the breakdown of glycogen and the increased production of glucose 6-phosphate, one of the intermediaries of the glycolysis/gluconeogenesis pathway (Constantin et al., 1995) (Figure 4). The balance between the resulting glycolysis and gluconeogenesis is also


DOI: 10.3109/17435390.2014.895437


5

Figure 2. Intermediary metabolism from lactate and pyruvate (10/1) substrates and PEPCK expression in C3A cells exposed to ZnO or TiO2 nanomaterial. C3A cells were submitted to nanomaterials (0–10 lg/cm2), for 4 hours, before metabolism measurements from 20 mM lactate, 2 mM pyruvate and 2 mM octanoate and 4 mM ammonium chloride, before metabolism measurements. (A) glucose production; (B) cytosolic redox potential *p50.0001 vs TiO2; and (C) PEPCK expression analysis was done by qRT-PCR, after incubation with NM or ZnCl. p50.001 overall for ZnO vs TiO2.

regulated by the availability of ATP. Hence, the large increase in both glycogen breakdown and glycolysis indicates that ZnO induces a drop of the cytosolic phosphate potential (the ratio of ATP to ADP and Pi) in C3A cells. Kao et al. (2012) have shown that repeated incubation of human lung epithelial cells with similar concentrations of ZnO nanoparticles resulted in an increase in both the cytosolic and mitochondrial concentration of Zn2+, to 150 and 10 nM, respectively. Zn2+ has previously been shown to strongly inhibit ATP synthesis by inhibiting both the mitochondrial electron transport chain and the Krebs cycle (for a review of the effects of Zn2+ on energy metabolism, see Dineley et al. (2003)). Therefore, despite the fact that, in this study, the incubation of C3A cells with ZnO was much shorter than in Kao’s studies, our results suggest similar effects on ATP concentration. Indeed, the measurement of the mitochondrial membrane potential (DWm) shows a strong depression, similar to that seen with CCCP. A drop of the DWm has previously been reported as one of the effects on mitochondrial function of Zn2+ (Dineley et al., 2003) and of ZnO nanoparticle exposure, in the same range of concentrations as used here (Jeng & Swanson, 2006), which supports our finding on glycogen breakdown. The fact that we show an increased glycolysis from endogenous substrates when the cells are submitted to ZnO seems to contradict the findings that Zn2+ inhibits glycolysis in neurons (Sheline et al., 2000). However, it is actually possible to have an increased enzyme velocity or pathway flux despite the inhibition of an enzyme, when there is a large increase in the enzyme substrate concentration due to the activation of an upstream

pathway. In this case, the increased glycogen breakdown would result in an increased glucose 6-phoshate concentration, resulting in an increased glycolysis flux despite the inhibition of the pathway by Zn2+. Furthermore, in their study, Sheline et al. suggested that the strong inhibition of glycolysis was due to the inability of the cells to regenerate NAD+ from NADH. In our case, the L/P ratio, a substitute marker for the NADH/NAD+ ratio (Williamson et al., 1967) was only moderately increased, showing that the capacity of hepatocytes to regenerate NAD+ was not radically affected. Because of the complex regulation of gluconeogenesis by catecholamines and the glucagon/insulin ratio – which are difficult to mimic in vitro, except using a perifusion system – further in vivo work on the effects of ZnO NM on gluconeogenesis would be particularly interesting. Furthermore, we were not able, in the short time allocated for this study, to determine the mechanism leading to the uptake of NM in hepatocytes, and whether endocytosis/pinocytosis was involved. This would also be an interesting research avenue. Effect of ZnO on hepatocyte glucose production In neurons, ZnO2 has already been shown to inhibit glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Sheline et al., 2000), resulting in the accumulation of one of the intermediaries of glycolysis, upstream of GAPDH, dihydroxyacetone phosphate (DHAP). Such accumulation of DHAP would lead to an increase in glucose production from glycogen stores, despite strong evidence of ATP depletion.

C. Filippi et al. c

3.4 4.20 1.24

The data present the ratio between the results obtained with 10 lg/cm2 and control cells, as well as the ratio between the ZnO and TiO2 effects. a p50.0001; bp50.01; and cp50.001 for ZnO vs TiO2. Most data shows no effect for TiO2 whilst ZnO increases metabolic pathway fluxes from 4.5- to 21-folds.

c

2.1 2.04 0.97

a

21.7 20.82 0.96 3.1 2.93 0.94

b a

6.0 4.45 0.74

a

11.2 10.20 0.91 6.5 4.88 0.74

a

ZnO TiO2 ZnO/TiO2 ZnO ZnO TiO2 ZnO/TiO2 ZnO ZnO/TiO2 ZnO TiO2

2G + L + P

TiO2

J Gluc

ZnO/TiO2

TiO2

ZnO

J L+P

L/P

ZnO/TiO2

TiO2

ZnO

ZnO/TiO2

TiO2

L/P J Gluc

LPON exogenous substrates Endogenous substrates

Table 2. Recap of the effects of ZnO and TiO2 on cell metabolism, from endogenous or LPON substrates, and PEPCK mRNA expression.


ZnO/TiO2


PEPCK expression

6

This probably could explain the increased JGlc in our cells from endogenous substrates. In our study, we also demonstrate that PEPCK expression is increased after ZnO exposure, resulting in a higher gluconeogenesis, most probably due to increased intracellular ROS production. Interestingly, this is similar to a cell culture model of NAFLD, induced by excess mitochondrial ROS production (Lockman et al., 2012). One could, therefore, hypothesise that long-term in vitro and, more importantly, in vivo exposure of hepatocytes to NMs may exacerbate intracellular ROS production and could lead to liver metabolic changes. On the other hand, since primary hepatocyte metabolism, including ROS scavenging capacity, has been shown to be different to HepG2/C3A cell metabolism, the extent of damage/metabolic variations generated by NM exposure may be different to that measured here. Exposure time considerations Four percent of ZnO became soluble in our system, after 4 hours, which corresponds to a Zn2+concentration of 60 lM. If the dissolution process is linear, it will only require 100 hours – or 4 days – to obtain total dissolution of the ZnO NM into Zn2+ ions. Though efficient faecal excretion of Zn2+ would take place at the same time (Baek et al., 2012) and limit the Zn2+ concentration increase, it would be particularly important to exactly monitor Zn2+ serum concentrations in in vivo chronic studies, in the light of previous work on the effects of Zn2+ on cell bioenergetics (for review see Dineley et al., 2003). Indeed, Sheline et al. described a time-dependent loss of cell ATP concentration and viability when neuronal cultures were exposed to concentrations of Zn2+ of 40 lM, whilst having glucose as sole energy substrate. They also showed that this effect was dramatically reduced by the addition of pyruvate. For hepatocytes, the presence of pyruvate would probably much better correspond to an in vivo situation, as the liver is responsible for the recycling of pyruvate produced by other organs. Thus, actual time-dependent serum concentrations of Zn2+ and liver metabolites should be considered for future in vivo studies. Trojan horse effects Our data show that only 4%, or 60 lM, of ZnO became soluble in our system. However, as the 10 lg/cm2 particle dose had an effect on PEPCK mRNA expression close to the corresponding dose of 100% dissolved Zn2+, or 1.23 mM, the effect observed could not be due solely to soluble ions in the medium, suggesting another mechanism, possibly a ‘‘Trojan horse’’ effect. Incubating cells with particles soluble only at low pH can result in a so-called ‘‘Trojan horse’’ effect in which particles deposited on the cell surface enter the cells through diffusion, endocytosis or pinocytosis, without any control for the final intracellular Zn2+ concentration, which then dissolve in acidic compartments inside the cells. If a high number of particles enter the cells, this leads to an intracellular concentration potentially much higher than if the cells had been solely submitted to the dissolved Zn2+concentration in the extracellular compartment (in our case 60 lM), when the entry of ions is controlled by specific channels. Previous measurement of the same ZnO nanoparticle solubility had shown that, after 24 hours at pH 5.5 in an artificial lysosomal fluid medium, the solubility of the particles was close to 100% (Cho et al., 2011). Therefore, we may have such a Trojan horse effect as far as ZnO is considered in our C3A cells. Effects of TiO2 TiO2 did not affect C3A cell metabolism and bioenergetics, in the range of concentrations and exposure time chosen. This result was


DOI: 10.3109/17435390.2014.895437


7

Figure 3. Detection of intracellular ROS formation in C3A cells submitted to NM. After incubation with NM or 5 mM H2O2 for 4 hours, C3A cells were rinsed, trypsinised and stained with DHE before being analysed by FACS. Typical FACS dot plot are presented for control cells or cells treated with H2O2, Ag, TiO2 or ZnO NM (all at 10 lg/cm2). Controls with unstained cells did not show any significant shift in signal due to potential interaction of intracellular NMs with the fluorescence measurement. The graph below shows the average data of n ¼ 3 experiments for a range of NM concentrations. *p50.0001 for TiO2 vs ZnO.

somewhat surprising seeing the high EPR signal that TiO2 produced on its own, a measurement of high endogenous ROS production. However, measurement of intracellular ROS production in cells exposed to TiO2 did not correlate with the EPR data. Various effects of TiO2 have been described. One thorough study, analysing 67 parameters of cell injury, inflammation or immune response, found no pathological changes induced by TiO2 after intravenous administration, despite accumulation and subsequent high TiO2 levels in the liver for over 28 days (Fabian et al., 2008). However, in an in vitro/ex vivo study of the effect of TiO2 on the brain (Long et al., 2007), TiO2 induced reactive species formation in the microglia, partly due to mitochondrial dysfunction, as well as strong effects on the mRNA expression of proteins involved in bioenergetics and the intermediary metabolism regulation. The discrepancy with our results may be due to variations of TiO2 NM composition as it can be found in two forms, rutile or anatase, with or without coating, that have extremely different physical and chemical properties. As seen in the different responses within brain cells, this may also be due to a lower sensitivity in hepatocytes or lower intracellular translocation in hepatocytes versus brain cells.

Conclusion Our study aimed to determine the effects of ZnO and TiO2 nanoparticles on hepatocyte intermediary metabolism. Using the C3A cell line and incubations time of 4 hours with concentrations of particles up to 10 lg/cm2, well below the particle EC50, we have shown that ZnO had a strong deleterious effect of mitochondrial function, resulting in a strong stimulation of glycogenolysis (Figure 4). Paradoxically, for a situation of energetic deficiency, the mitochondrial function decrease was concomitant with a large increase in the gluconeogenic flux from LP, resulting from the over-expression of PEPCK and potentially concomitant inhibition of glycolysis. Part of these effects, similar to effects of Zn2+ on energy metabolism, are probably due to the solubility of the particles and release of Zn2+ inside the cells. TiO2 nanoparticles did not exhibit any deleterious effects, at the range of concentrations used in our study, confirming previous data presenting TiO2 as a low toxicity NM. Air-borne particulate matter, which shares some of the characteristics of ZnO particles, such as the capacity to mediate an oxidative stress, to induce lung inflammation and to be


8

C. Filippi et al.

Figure 4. Schematic representation of the effects of ZnO on C3A cell metabolism from endogenous glycogen or lactate and pyruvate. (i) The incubation of C3A cell with ZnO leads to a drop in the DWm (¶), inducing a drop in mitochondrial ATP synthesis (•) and cytosolic ATP. This releases the inhibition of ATP on glycogen metabolism and provokes a large increase in glycogen breakdown (‚) and glucose 6-phosphate production. In a classical enzyme kinetics effect, this increases the glucose production as well as the glycolysis, further enhanced by the ATP drop. (ii) The gluconeogenesis from lactate and pyruvate was largely enhanced after ZnO exposure of C3A cells, due to the increase in the expression of the PEPCK („), the enzyme controlling this metabolic pathway, probably following the rise of intracellular ROS production (”). According to Sheline et al. GAPDH is also inhibited by intracellular Zn2+, increasing DHAP concentrations (») and further increasing the flux of glucose production (…).

cytotoxic at high concentrations, has recently been shown to exacerbate NAFLD, when instilled in the lungs in animal models. Taken together with our data, this seems to be a strong motivation for the study of ZnO effects – and other metal oxide particles – on liver metabolism, at dosages where there are little or no observable cytotoxic effects, both in control primary cells and in cell culture models of NAFLD. Though we have shown very strong effects of ZnO in an acute exposure situation on isolated cells, in vivo Zn2+ has been shown to be efficiently eliminated from the blood stream. In order to assess ZnO NM effect on liver metabolism, animal studies in chronic conditions should also probably be undertaken.

Declaration of interest Conflict of interest: none. Financial Support: Sir Jules Thorn Trust ‘‘Seed Corn Fund’’ to Ce´line Filippi.

References Nanotechnology consumer products inventory. Available at: http:// www.nanotechproject.org/inventories/consumer/. Accessed on 5 January 2012. Baek M, Chung HE, Yu J, Lee JA, Kim TH, Oh JM, et al. 2012. Pharmacokinetics, tissue distribution, and excretion of zinc oxide nanoparticles. Int J Nanomedicine 7:3081–97. Bergmeyer HU, Gawehn ICWK, ed. 1974. Methods of Enzymatic Analysis. Weinheim: Verlag Chemie; New York: Academic Press.


Brayner R, Ferrari-Iliou R, Brivois N, Djediat S, Benedetti MF, Fie´vet F. 2006. Toxicological impact studies based on Escherichia coli bacteria in ultrafine ZnO nanoparticles colloidal medium. Nano Lett 6:866–70. Cho WS, Duffin R, Howie SEM, Scotton CJ, Wallace WAH, Macnee W, et al. 2011. Progressive severe lung injury by zinc oxide nanoparticles; the role of Zn2+ dissolution inside lysosomes. Part Fibre Toxicol 8:27. Constantin J, Ishii-Iwamoto E, Suzuki-Kemmelmeier F, Yamamoto N, Bracht A. 1995. Bivascular liver perfusion in the anterograde and retrograde modes: zonation of the response to inhibitors of oxidative phosphorylation. Cell Biochem Funct 13:201–9. Dineley KE, Votyakova TV, Reynolds IJ. 2003. Zinc inhibition of cellular energy production: implications for mitochondria and neurodegeneration. J Neurochem 85:563–70. Donaldson K, Tran L, Jimenez LA, Duffin R, Newby DE, Mills N, et al. 2005. Combustion-derived nanoparticles: a review of their toxicology following inhalation exposure. Part Fibre Toxicol 2:10. Fabian E, Landsiedel R, Ma-Hock L, Wiench K, Wohlleben W, van Ravenzwaay B. 2008. Tissue distribution and toxicity of intravenously administered titanium dioxide nanoparticles in rats. Arch Toxicol 82: 151–7. Groen A, Vervoorn R, Van D, Tager J. 1983. Control of gluconeogenesis in rat liver cells. I. Kinetics of the individual enzymes and the effect of glucagon. J Biol Chem 258:14346–53. Hussain SM, Hess KL, Gearhart JM, Geiss KT, Schlager JJ. 2005. In vitro toxicity of nanoparticles in BRL 3A rat liver cells. Toxicol In Vitro 19: 975–83. Ito Y, Oumi S, Nagasawa T, Nishizawa N. 2006. Oxidative stress induces phosphoenolpyruvate carboxykinase expression in H4IIE cells. Biosci Biotechnol Biochem 70:2191–8. Jani P, Halbert G, Langridge J, Florence A. 1990. Nanoparticle uptake by the rat gastrointestinal mucosa: quantitation and particle size dependency. J Pharm Pharmacol 42:821–6. Jeng HA, Swanson J. 2006. Toxicity of metal oxide nanoparticles in mammalian cells. J Environ Sci Health A Tox Hazard Subst Environ Eng 41:2699–711. Jones N, Ray B, Ranjit KT, Manna AC. 2008. Antibacterial activity of ZnO nanoparticle suspensions on a broad spectrum of microorganisms. FEMS Microbiol Lett 279:71–6. Kamat JP, Devasagayam TP, Priyadarsini KI, Mohan H, Mittal JP. 1998. Oxidative damage induced by the fullerene C60 on photosensitization in rat liver microsomes. Chem Biol Interact 114:145–59. Kao Y, Chen Y, Cheng T, Chiung Y, Liu P. 2012. Zinc oxide nanoparticles interfere with zinc ion homeostasis to cause cytotoxicity. Toxicol Sci 125:462–72. Kermanizadeh A, Gaiser B, Hutchison G, Stone V. 2012. An in vitro liver model – assessing oxidative stress and genotoxicity following exposure of hepatocytes to a panel of engineered nanomaterials. Part Fibre Toxicol. 9:28. Kermanizadeh A, Pojana G, Gaiser B, Birkedal R, Bilanicova´ D, Wallin H, et al. 2013. In vitro assessment of engineered nanomaterials using a hepatocyte cell line: cytotoxicity, pro-inflammatory cytokines and functional markers. Nanotoxicology 7:301–13. Khandoga A, Stampfl A, Takenaka S, Schulz H, Radykewicz R, Kreyling W, Krombach F. 2004. Ultrafine particles exert prothrombotic but not inflammatory effects on the hepatic microcirculation in healthy mice in vivo. Circulation 109:1320–5. Klingshirn C. 2007. ZnO: material, physics and applications. Chemphyschem 8:782–803. Koukkunen H, Penttila¨ K, Kemppainen A, Halinen M, Penttila I, Rantanen T, Pyo¨ra¨la¨ K. 2001. C-reactive protein, fibrinogen, interleukin-6 and tumour necrosis factor-alpha in the prognostic classification of unstable angina pectoris. Ann Med 33:37–47. Kreyling WG, Semmler-Behnke M, Seitz J, Scymczak W, Wenk A, Mayer P, et al. 2009. Size dependence of the translocation of inhaled iridium and carbon nanoparticle aggregates from the lung of rats to the blood and secondary target organs. Inhal Toxicol 21:55–60. Lee SJ, Jin Y, Yoon HY, Choi BO, Kim HC, Oh YK, et al. 2005. Ciclopirox protects mitochondria from hydrogen peroxide toxicity. Br J Pharmacol 145:469–76. Leverve XM, Fontaine E, Putod-Paramelle F, Rigoulet M. 1994. Decrease in cytosolic ATP/ADP ratio and activation of pyruvate kinase after in vitro addition of almitrine in hepatocytes isolated from fasted rats. Eur J Biochem 224:967–74. Li C, Shen C, Cheng Y, Huang S, Wu C, Kao C, et al. 2012. Organ biodistribution, clearance, and genotoxicity of orally administered zinc oxide nanoparticles in mice. Nanotoxicology 6:746–56.


DOI: 10.3109/17435390.2014.895437


Lockman K, Baren J, Pemberton C, Baghdadi H, Burgess K, Plevris-Papaioannou N, et al. 2012. Oxidative stress rather than triglyceride accumulation is a determinant of mitochondrial dysfunction in in vitro models of hepatic cellular steatosis. Liver Int 32:1079–92. Long T, Tajuba J, Sama P, Saleh N, Swartz C, Parker J, et al. 2007. Nanosize titanium dioxide stimulates reactive oxygen species in brain microglia and damages neurons in vitro. Environ Health Perspect 115: 1631–7. Maheshwari A, Thuluvath PJ. 2006. Cryptogenic cirrhosis and NAFLD: are they related? Am J Gastroenterol 101:664–8. Miller MR, Borthwick SJ, Shaw CA, McLean SG, McClure D, Mills NL, et al. 2009. Direct impairment of vascular function by diesel exhaust particulate through reduced bioavailability of endothelium-derived nitric oxide induced by superoxide free radicals. Environ Health Perspect 117:611–16. Miller R, Lenihan H, Muller E, Tseng N, Hanna S, Keller A. 2010. Impacts of metal oxide nanoparticles on marine phytoplankton. Environ Sci Technol 44:7329–34. Monn C, Becker S. 1999. Cytotoxicity and induction of proinflammatory cytokines from human monocytes exposed to fine (PM2.5) and coarse particles (PM10-2.5) in outdoor and indoor air. Toxicol Appl Pharmacol 155:245–52. Oberdo¨rster G, Maynard A, Donaldson K, Castranova V, Fitzpatrick J, Ausman K, et al. 2005. Principles for characterizing the potential human health effects from exposure to nanomaterials: elements of a screening strategy. Part Fibre Toxicol 2:8. Oberdo¨rster G, Sharp Z, Atudorei V, Elder A, Gelein R, Lunts A, et al. 2002. Extrapulmonary translocation of ultrafine carbon particles following whole-body inhalation exposure of rats. J Toxicol Environ Health A 65:1531–43. Pessayre D. 2007. Role of mitochondria in non-alcoholic fatty liver disease. J Gastroenterol Hepatol 1:S20–7. Pujalte´ I, Passagne I, Brouillaud B, Tre´guer M, Durand E, Ohayon-Courte`s C, L’Azou B. 2011. Cytotoxicity and oxidative stress induced by different metallic nanoparticles on human kidney cells. Part Fibre Toxicol 8:10. Reed R, Ladner D, Higgins C, Westerhoff P, Ranville J. 2012. Solubility of nano-zinc oxide in environmentally and biologically important matrices. Environ Toxicol Chem 31:93–9. Rognstad R. 1979. Rate-limiting steps in metabolic pathways. J Biol Chem 254:1875–8. Saquib Q, Al-Khedhairy A, Siddiqui M, Abou-Tarboush F, Azam A, Musarrat J. 2012. Titanium dioxide nanoparticles induced cytotoxicity, oxidative stress and DNA damage in human amnion epithelial (WISH) cells. Toxicol In Vitro 26:351–61. Sharma V, Anderson D, Dhawan A. 2011. Zinc oxide nanoparticles induce oxidative stress and genotoxicity in human liver cells (HepG2). J Biomed Nanotechnol 7:98–9.

9

Sharma V, Singh P, Pandey A, Dhawan A. 2012. Induction of oxidative stress, DNA damage and apoptosis in mouse liver after sub-acute oral exposure to zinc oxide nanoparticles. Mutat Res 745:84–91. Sheline C, Behrens M, Choi D. 2000. Zinc-induced cortical neuronal death: contribution of energy failure attributable to loss of NAD(+) and inhibition of glycolysis. J Neurosci 20:3139–46. Stone V, Donaldson K. 2006. Nanotoxicology: signs of stress. Nat Nanotechnol 1:23–4. Stone V, Johnston H, Clift MJD. 2007. Air pollution, ultrafine and nanoparticle toxicology: cellular and molecular interactions. IEEE Trans Nanobioscience 6:331–40. Sun L, Li Y, Liu X, Jin M, Zhang L, Du Z, et al. 2011. Cytotoxicity and mitochondrial damage caused by silica nanoparticles. Toxicol In Vitro 25:1619–29. Sutherland C, Tebbey PW, Granner DK. 1997. Oxidative and chemical stress mimic insulin by selectively inhibiting the expression of phosphoenolpyruvate carboxykinase in hepatoma cells. Diabetes 46: 17–22. Sycheva L, Zhurkov V, Iurchenko V, Daugel-Dauge N, Kovalenko M, Krivtsova E, Durnev A. 2011. Investigation of genotoxic and cytotoxic effects of micro- and nanosized titanium dioxide in six organs of mice in vivo. Mutat Res 726:8–14. Tan HH, Fiel MI, Sun Q, Guo J, Gordon RE, Chen LC, et al. 2009. Kupffer cell activation by ambient air particulate matter exposure may exacerbate non-alcoholic fatty liver disease. J Immunotoxicol 6: 266–75. Turrens J. 2003. Mitochondrial formation of reactive oxygen species. J Physiol 552:335–44. Wang Y, Aker W, Hwang H, Yedjou C, Yu H, Tchounwou P. 2011. A study of the mechanism of in vitro cytotoxicity of metal oxide nanoparticles using catfish primary hepatocytes and human HepG2 cells. Sci Total Environ 409:4753–62. Weir A, Westerhoff P, Fabricius L, Hristovski K, von G. 2012. Titanium dioxide nanoparticles in food and personal care products. Environ Sci Technol 46:2242–50. Williamson DH, Lund P, Krebs HA. 1967. The redox state of free nicotinamide-adenine dinucleotide in the cytoplasm and mitochondria of rat liver. Biochem J 103:514–27. Xia T, Kovochich M, Liong M, Ma¨dler L, Gilbert B, Shi H, et al. 2008. Comparison of the mechanism of toxicity of zinc oxide and cerium oxide nanoparticles based on dissolution and oxidative stress properties. ACS Nano 2:2121–34. Xia T, Korge P, Weiss JN, Li N, Venkatesen MI, Sioutas C, Nel A. 2004. Quinones and aromatic chemical compounds in particulate matter induce mitochondrial dysfunction: implications for ultrafine particle toxicity. Environ Health Perspect 112:1347–58.

Investigation on likely effects of Ag, TiO2, and ZnO nanoparticles on sewage treatment.

Impact of particle size and light exposure on the effects of TiO2 nanoparticles on Caenorhabditis elegans.

The effect of Fe2O3 and ZnO nanoparticles on cytotoxicity and glucose metabolism in lung epithelial cells.

Enhanced Photoelectrochemical Activity of ZnO-Coated TiO2 Nanotubes and Its Dependence on ZnO Coating Thickness.

Effects of acute systemic administration of TiO2, ZnO, SiO2, and Ag nanoparticles on hemodynamics, hemostasis and leukocyte recruitment.

Impact of manufactured TiO2 nanoparticles on planktonic and sessile bacterial communities.

Characterization of TiO2 and ZnO nanoparticles and their applications in photocatalytic degradation of azodyes.

Preparation and Characterization of ZnO Nanoparticles Supported on Amorphous SiO₂.

Effects of TiO2 nanoparticles on nutrition metabolism in silkworm fat body.

Synthesis of TiO2 nanotubes with ZnO nanoparticles to achieve antibacterial properties and stem cell compatibility.

Toxicological mode of action of ZnO nanoparticles: Impact on immune cells.

Epigenetic modulation upon exposure of lung fibroblasts to TiO2 and ZnO nanoparticles: alterations in DNA methylation.

Arsenite Effects on Mitochondrial Bioenergetics in Human and Mouse Primary Hepatocytes Follow a Nonlinear Dose Response.

The impact of cryopreservation on human peripheral blood leucocyte bioenergetics.

Oxidative Stress and Nano-Toxicity Induced by TiO2 and ZnO on WAG Cell Line.

Effect of humic acid on photocatalytic activity of ZnO nanoparticles.

Effect of exopolysaccharides on photocatalytic activity of ZnO nanoparticles.

Tenth anniversary of Expert Opinion on Drug Metabolism & Toxicology.

Effect of TiO2 nanoparticles on the reproduction of silkworm.

The effect of ZnO nanoparticles on liver function in rats.

Cytotoxic effects of ZnO nanoparticles on mouse testicular cells.

Enhanced antimicrobial and anticancer properties of ZnO and TiO2 nanocomposites.

Mechanism of photocatalytic water oxidation on small TiO2 nanoparticles.

Impact of TiO₂ and ZnO nanoparticles at predicted environmentally relevant concentrations on ammonia-oxidizing bacteria cultures under ammonia oxidation.